HAL S. SCOTT [*]
In fully internationalized securities markets, issuers in public primary markets should be able to issue securities to investors worldwide using one set of optimal distribution procedures and disclosure documents, and subject to one set of liability standards and enforcement remedies. Rules in which the benefits outweigh the costs are optimal. Optimal standardized issuance ("OSI") across borders would reduce the costs of issuing securities that are in international demand, a benefit that would be shared by both issuers and investors. Also, OSI would result in more perfect competition in the issuing market for such securities (just as in the goods market) and a more efficient allocation of capital worldwide. OSI is only relevant when there is sufficient investor demand for a public international distribution. This article is concerned only with those cases. The significant growth of registered offerings of foreign issuers in the United States, from seventy-eight in 1990 to 526 (with a total value of $145.9 bi llion) in 1998, is a strong indication of such investor demand. 
This article focuses on public primary markets rather than public secondary markets. While the two markets are related--publicly distributed securities are invariably listed and traded in public secondary markets--they are not the same. Public securities are initially distributed through an underwriting process where securities are sold to investors by the issuer through underwriters; thereafter, the securities are traded in the secondary market. Primary market distributions play a crucial part in the allocation of capital. Investors purchasing in primary markets, as opposed to secondary markets, cannot necessarily rely on prices set in deep liquid markets where rational expectations of the value of the securities have been incorporated into the price.  Additionally, public primary markets involve purchases by individual investors; indeed, issuers make such distributions in order to reach these investors. While initial public offerings ("IPO"s)--an important part of the public primary market--are heavily dominated by institutional investors, individual investors participate significantly in these markets, as well.
This article assumes that there is no conflict between having standardized and optimal disclosure rules. Merritt Fox, to justify his issuer nationality approach, argues to the contrary when he raises the possibility that the socially optimal level of disclosure may differ for different countries.  He argues, in effect, that the optimal level of disclosure in Germany and Japan, where capital has been traditionally allocated mainly through banking markets, may be lower than in the United States, where securities markets have played a more important role.  However, in major markets, including Germany and Japan, securities markets currently play an important enough role that those countries care about how they allocate capital. Securities markets in these countries, as compared to banking markets, are gaining importance. Further, even if one should be more tolerant of less disclosure in Germany, such a rule would only affect allocation of capital within Germany. Such tolerance would not be appropriate in a n international context where the concern is with allocation among issuers of different countries. If Deutsche Telekom raises significant capital outside Germany, we should be concerned with choosing the right set of rules for the world, not just Germany.
Part II of this article discusses the current obstacles to achieving OSI. It argues that the imposition of U.S. rules for disclosure, distribution, and enforcement, within and to some extent outside the United States, impedes the achievement of OSI. It also argues that the solution is not merely to abolish mandated disclosure. The article then proceeds to examine options for dealing with the problem. Parts III and IV, respectively, examine and then reject harmonization and various versions of mutual recognition as solutions. Part V examines the structure of U.S. Regulation S and restrictions on use of the Internet. Part VI proposes that relaxation of these rules can substantially contribute to establishing OSI for the issuance of foreign securities. Part VII is a conclusion.
THE OPTIMALITY OF U.S. DISCLOSURE RULES
Currently, OSI is not achievable in public securities markets, largely because the United States conditions issuance in its territory--and to some significant extent to U.S. investors outside its territory--on compliance with its unique set of distribution procedures, disclosure documents, and enforcement rules. It is well known that this practice has deterred primary issuers from distributing securities in U.S. markets and to U.S. investors. Standardized issuance ("SI") can be achieved if issuers are willing to abide by U.S. rules, because these rules are tougher in most material respects than those in other jurisdictions. But it is unlikely that these rules are optimal. Because the United States is the predominant source of worldwide capital, U.S. rules have either substantially fractionated international public securities markets by precluding U.S. issues, or resulted in non-optimal standardized distributions.
A. The Market Test of Optimality
The fact that the international unregulated private placement market does not always demand U.S. public market documentation reveals that the level of disclosure required by the United States is not optimal. Instead, the international market generally has adopted what market participants call "international" documentation or rules, developed by issuers, underwriters, and institutional investors.  The United States justifies its more complicated disclosure rules by claiming that they are for the benefit or protection of the less sophisticated investor. It is unlikely, however, that relatively unsophisticated individual investors can make use of more or different information than sophisticated investors; indeed, because of their lack of sophistication, they can make less use of such information. To the extent that the rationale for disclosure levels is the desirability of accurate market prices, sophisticated investors have just as much incentive, and probably more, in having such accuracy.
It is more plausible to argue that individual investors need to have information presented to them in a different manner than do sophisticated investors. This idea is central to the SEC's advancement of the "Plain English" requirement.  Sophisticated investors, it is argued, do not need to have things spelled out because they bring more background to their reading. Yet this argument is faulty as well, because most individual investors do not read prospectuses; the information in the prospectuses is predigested for them by analysts or advisors. Those who do read prospectuses are likely to be fairly sophisticated, even if they would not qualify as private placement investors.  In any event, distinguishing between style and substance, we can use the substance of disclosure in the private market as a benchmark. It is important to emphasize that private market disclosure standards arise out of commercial practice and not law. Indeed, one cannot go anywhere to look them up--probably the best one could do is t o reverse engineer the rules from the actual forms lawyers use to create their documentation.
Further research is necessary to determine how closely the international standard matches existing U.S. rules for public offerings. One clear difference does exist: There is no U.S. GAAP reconciliation under the international standard. If the United States were to accept the International Accounting Standards Committee ("IASC") standards as an alternative to U.S. standards, as now seems quite probable, this difference may become immaterial. Apart from accounting, it appears that the international standard may be quite close to U.S. rules for some offerings, particularly IPOs. But for secondary offerings  of widely traded companies--which regularly report material information to the market--there may be a significant difference between international standards and U.S. rules. For example, in a recent private equity offering for Hong Kong and Shanghai Bank, documentation was minimal because the market believed it already had enough information. If private market disclosure based on actual practices is used a s a basis for determining the adequacy of public disclosure, the level of required disclosure would depend on various characteristics of the issue--for example, debt versus equity, IPO versus secondary, and big versus small--as well as the nature of the issuer--for example, reporting versus non-reporting and regulated (for example, banks) versus non-regulated.
The solution to different distribution and enforcement rules is much less clear. It is entirely rational to have different distribution rules for institutional and individual investors. For example, one may have legitimate concerns that individual investors could be more influenced than institutional investors to buy securities based on underwriter research reports disseminated during the distribution period. It also may be entirely rational to have different enforcement rules for public and private markets. For example, the feasibility of enforcement by individuals in public markets may depend on the prospect of damages in a class action suit, while in private markets, the necessity of repeat dealings may adequately deter violations.  Unfortunately, there is no reliable private market for distribution rules or enforcement to use as a benchmark for the optimality of rules.
An issuer can generally deal with diverse national distribution requirements in three ways: (1) abstain entirely from issuing in the U.S. market; (2) distribute in foreign markets under foreign rules and in the U.S. market under U.S. rules, which might result in the distribution being later in the United States than abroad; or (3) refrain from issuing marketing materials, other than a U.S.-compliant prospectus, in any market until the registration statement is effective in the United States. Option 1 clearly results in segmented markets, as no securities are issued in the United States. Option 2 also results in segmentation, because price competition introduced by U.S. distributions is not available during earlier distributions in other countries. Option 3 achieves a standardized distribution, but at the cost of deferring the distribution until the U.S. effective registration date. Option 3 could increase investor cost if deferred marketing increases issuer distribution cost, and may not be justified if it d ecreases investor welfare overall. This is a general problem with any system of mandated harmonized rules, as discussed in Part III below.
The SEC is largely responsible for engineering the current state of affairs. It has acted from the best of motives--the protection of U.S. investors. It seeks generally to ensure that investors are well-informed before making investment decisions and believes that rigorous disclosure is necessary to protect the reputation of U.S. capital markets. The SEC believes in the virtues of standardized primary issuance markets, but insists that U.S. law set the standard.
In my view, the SEC has pursued a limited vision of investor welfare. In the desire to protect individual investors (visions of widows and orphans spring to mind), the SEC may have sacrificed investor welfare as a whole. While some number of investors may make better investment decisions as a result of SEC rules, investor welfare as a whole is decreased by the inability to achieve standardized rules. Lack of standardization imposes higher costs on issuers, which are passed on to investors through higher prices for capital. As in other areas of consumer protection, the benefits to particular consumers should be weighed against the costs to consumers as a whole.
B. Claims That Disclosure is Unimportant to Investor Welfare
Some authors have taken the extreme view that investor welfare is largely unaffected by the level of disclosure. They argue that investors will take the level of disclosure into account when purchasing securities. Investors will discount the price of securities for lack of disclosure because securities with less disclosure will be more risky than those with more disclosure.  This argument is entirely unconvincing, particularly in primary markets. First, there is generally no way for an investor, sophisticated or otherwise, to know that he is missing information and to attach an appropriate discount to that risk.  For example, an investor has no way of knowing that an issuer failed to disclose material information, such as a design flaw.  Even if the investor knows that the issuer failed to disclose a particular type of information, such as segmented line of business information, the investor does not know that this means the issuer is more risky. The issuer may have chosen not to disclose such in formation because disclosure would aid competitors, rather than because it would show that its earnings were undiversified. One could try to finesse this basic problem by arguing that the price of a security reflects whatever judgments, however imperfect, the market makes about the adequacy of disclosure. So while the discounting may not be perfect--no discount ever is--it does occur.
Second, the discounting argument only works if a market generally has adequate disclosure, and that the price of an individual security based on inadequate disclosure can be discounted; that is, that there is a market baseline against which to measure the discount. However, if the market provides generally inadequate information, such discounting becomes impossible. In addition, whatever persuasiveness this argument may have in secondary markets, it fails in primary markets because the price of the security is not generally set by the market, but by the issuer and the underwriter. Furthermore, the discounting argument ignores the impact of such discounts on the capital formation process. Due to the riskiness of markets without optimal disclosure rules, there will be underinvestment in such markets.  At the macro level, worthwhile projects may go without funding; at the micro level, investor preferences for security investments with less risk will go unsatisfied. In theory, markets suffering from underinv estment due to sub-optimal disclosure should, in the long term, increase disclosure requirements. But increasing disclosure requirements leaves no protection for investors in the short term. Moreover, countries may fail to adopt optimal disclosure rules because of regulatory capture by influential issuers or inadequate resources.
Third, these discounting arguments ignore the problem of over-disclosure. As argued above, if issuers are required to disclose information where the costs of providing the information exceed the benefits to investors, investor welfare is decreased. Investors cannot avoid this problem by simply refusing to buy stocks issued in markets demanding too much disclosure if such markets are sizable, like those of the United States, and attractive on dimensions other than disclosure. 
A cousin of the discounting argument is the contention that the level of disclosure does not matter to investors who have diversified portfolios. As with the discount argument, this contention has very limited applicability to the primary market, as diversified portfolios would not usually contain new issues (particularly not IPOs). Further, there are basic problems with the diversification argument even for the secondary market.
The term "diversification" in this context is not very clear. The conventional notion of a portfolio diversified for the risks of a specific issuer--for example by type of business and country of operations--requires disclosure to obtain the information necessary to diversify if one is picking stocks (as opposed to buying index funds). In constructing a diversified portfolio, one seeks lack of co-variation in the price of the securities. This could be done mechanically based only on some period of past price performance, without knowing anything about the issuers of securities. But as we know, past is not always, and perhaps not even usually, prologue. Suppose, for example, that some industries had fixed the Y2K problem and others had not and could not. The past …